903group6 / pizza

Kaggle Competition

Random Acts of Pizza

The included code in this project is to produce predictions for the Kaggle
"Random acts of pizza" competition:
(https://www.kaggle.com/c/random-acts-of-pizza).

Usage: in a command line run "python main.py"

Requirements:

• scikit learn version 0.14.1 or later
• nosetests
• nltk with latest data packages (see: see http://stackoverflow.com/questions/17672423/http-proxy-authentification-error-for-nltk-download for steps to update path if necessary)

main.py(inputs)

In this file, you can update the following parameters: 

• Training and Test File locations
• Paths to output for models, evaluations results and kaggle submission file
• Name of your model
• Additional features to include beyond the baseline features
• MOST IMPORTANT: You can choose your classifier

Outputs

With these parameters specified, the program will output to the path:

• CSV file for submission to Kaggle

• Model Evaluation Files for Review

• Binary files to predict results for new Reddit Posts

  For full details of the system and results, please see the included report on our system.

  Note Also: This system and code can be adapted for similar problems

Pipeline

**"import raop pipeline"**  This module contains functions to preprocess data
and extract features.  Check out doc strings and code for details.